ATS Resume Guide for IT Professionals (2026): What Really Works

An ATS-friendly IT resume is a plain, single-column document with standard headings, contact details in the body, full job titles and the exact skill names from the job post, saved as a text-based DOCX or PDF. That's most of it. The rest is writing bullet points that a recruiter wants to read once the system has turned your file into a searchable profile.

A lot of ATS advice online is fear marketing. The line you've probably seen, that "75% of resumes are rejected by ATS before a human sees them", has no published study behind it. This guide separates what the big platforms actually do, based on their own documentation, from what resume-tool vendors claim. It covers how parsing works, what really rejects you automatically, formatting, keywords, bullet points, the skills section and how to test your file before you apply. Sources were checked on 25 September 2026.

Key takeaways

• The 75% figure is unverified. It's widely traced to a 2012 sales pitch from a resume-optimization company, Preptel, and no methodology was ever published. Treat it as a myth.
• Knockout questions reject people, not resume layouts. Greenhouse's auto-reject, for example, works from your answers to yes/no or multiple-choice questions on the application form.
• Parsing failures are real but fixable. Greenhouse lists tables, columns, text boxes, graphics, headers and footers, and files over 2.5 MB as common reasons a resume won't parse.
• AI match scores exist, and mostly sort rather than decide. Greenhouse's documentation describes Talent Matching as scoring applications for recruiters to review, not as an automatic rejection tool.
• Keywords still matter, because recruiters search the parsed profile. Use the job post's exact terms, but only for skills you can defend in an interview.

The "75% of resumes are rejected by ATS" claim is a myth

The exact percentage even changes depending on who repeats it. Investigations such as this one tracing the figure to its source and HiringThing's ATS myths piece point back to Preptel, a company that sold resume optimization and shut down in August 2013. No survey, sample size or method was ever released. A statistic without a source isn't a statistic, so I've dropped it from this guide.

What evidence there is points the other way. In a small Enhancv study, 25 US recruiters were interviewed in September and October 2025. 23 of them (92%) said their system doesn't automatically reject resumes on formatting, content or design. Most relied on knockout filters for eligibility questions such as work authorization or required certifications. It's a small sample from a resume-builder company, so read it as a signal, but it matches the vendors' own documentation.

So why do so many applications vanish? Mostly volume. Popular roles attract more applications than a recruiter can read closely, and someone reviewing them in order may stop long before reaching yours. The fixes there are different: apply early, tailor the resume, and get a referral where you can.

How an ATS actually processes your resume

An applicant tracking system is a database and workflow tool for recruiters. When you upload a resume, a parser extracts text and tries to map it to fields: name, email, phone, job titles, employers, dates, schools and skills. iCIMS describes parsing as technology that "extracts data, usually from a Word or PDF document, and converts it in a structured way, often self-completing a form."

Recruiters filter and search that structured profile. Your original file normally stays attached, but the parse decides whether you show up when someone searches for "Intune" or "ITIL 4".

Platform behavior, from vendor documentation

Other systems, such as Oracle Taleo or SAP SuccessFactors, follow the same broad model, and every employer configures its own. Nobody outside that company can tell you exactly how a given job is set up.

What actually rejects applications automatically

Knockout questions. A job form asks "Do you have the right to work in the UK?" or "Do you hold a CCNA?", and a "No" to a required item can move you to rejected without anyone reading your resume. The Enhancv recruiters gave a bachelor's degree question as their example.

Two habits help here:

• Read every question twice. "Five or more years of experience with Azure" is a different question from "five or more years in IT".
• Answer honestly. Saying yes to a certification you don't hold gets you past the form and fails you at the background or reference check, which is worse.

Formatting that parses cleanly

Greenhouse's list of parsing problems is the most specific vendor guidance I've found, and it's worth taking literally. It names spaced-out letters, graphics, photos, word art, image-format resumes, tables, headers and footers, contact details in text boxes, columned layouts and unclear sections. Two items are easy to miss: abbreviated job titles such as "Sr. Account Exec", and company names without a formal identifier such as Inc., LLC or Ltd.

For IT people in Bangladesh, that last point is a small, free fix. Write "Grameenphone Ltd." rather than just "Grameenphone", and "Senior Systems Administrator" rather than "Sr. Sys Admin".

Do and don't

DOCX or PDF? Both are on Greenhouse's supported list, so follow the job post. If it doesn't say, a PDF exported from Word or Google Docs is fine as long as you can select and copy the text in a PDF reader.

Keywords: match the job post, not a generic list

Recruiters search parsed profiles with the words they care about, and AI match scores compare your content against the job. So the practical rule is simple: use the posting's vocabulary for the things you've actually done.

1. Highlight every tool, platform, certification and method in the job post, then keep only the ones you've really used.
2. Use the employer's spelling. If they write "Microsoft Intune", don't write only "MEM". If they write "Jira Service Management", don't write "JSM".
3. Put each keyword where it makes sense: the Skills section and at least one bullet point showing you used it.

IT manager and analyst posts usually mix technical terms with words like budgeting, vendor management or stakeholder reporting. For the management side, IT management best practices in Bangladesh covers the vocabulary hiring managers use. For analyst roles, the methods in system analysis techniques for modern IT professionals give you accurate terms to describe your work.

Don't game it. Jobscan, which sells a resume-matching tool, recommends aiming for about an 80% match rate in its own tool and warns that "there is such thing as an over-optimized resume." Keep in mind that's Jobscan's score, not the one the employer's system calculates.

Bullet points that work for parsers and people

The parser needs your title, employer and dates on a clear line. Everything below that is for a human who wants to know what changed because you were there. A reliable pattern is action + what you worked on + measurable result.

Weak bullets describe a job description:

• "Responsible for managing the IT team"
• "Worked on network infrastructure projects"

Stronger bullets follow a template you fill with your own real numbers:

• "Led a [team size]-person IT team covering helpdesk, infrastructure and security; cut average ticket resolution time from [before] to [after] by introducing ITIL-based incident and problem management in ServiceNow."
• "Migrated [number] on-premises servers to Microsoft Azure, reducing annual hosting costs by [amount in BDT or USD]."
• "Gathered requirements from [number] departments for an ERP rollout and delivered it [on time / under budget by X%]."

No hard number? Use scale: users supported, sites, endpoints or the budget you controlled. Don't invent figures. Interviewers ask "how did you measure that?", and a vague answer undoes the bullet. Cloud work reads better when you name the service precisely; this comparison of AWS, Azure and GCP can help you get the terms right.

The skills section: group it, keep it plain

Plain text, one group per line, no tables or icons:

SKILLS
Cloud: Microsoft Azure, AWS, Cloudflare Workers
Networking: Cisco IOS, Fortinet FortiGate, VLANs, site-to-site VPN, DNS, DHCP
Security: Microsoft Defender for Endpoint, Wazuh, ISO/IEC 27001
Systems: Windows Server, Active Directory, Microsoft Intune, VMware ESXi
Databases: Microsoft SQL Server, MySQL, PostgreSQL
Scripting: Python, PowerShell, Bash
ITSM: ServiceNow, Jira Service Management, ITIL 4

CERTIFICATIONS
Microsoft Certified: Azure Administrator Associate
ITIL 4 Foundation
CompTIA Security+

LANGUAGES
Bengali (native), English (professional)

That's an example layout, not a list to copy. Include only what you'd be comfortable being quizzed on. If AI tools are part of your day job, name the specific ones and what you use them for; AI tools for IT managers has examples of how to describe that work concretely.

Employment gaps and career changes

There's no vendor documentation I could find saying these systems automatically flag date gaps, so I won't claim they do. A person reading your work history will notice one, though. A single line in the right place handles it: "Career break: completed AZ-104 and built a home lab for Intune and Defender testing" is better than an unexplained blank year.

Moving into management? Put the management work you did under a technical title, such as leading a team, owning a budget or running vendor evaluations, at the top of that role.

Test your resume before you apply

1. The copy-paste test. Open your PDF, select all, and paste into a plain text editor. If the order is scrambled, lines merge, or your name and email are missing, a parser will struggle too.
2. The file-size check. Keep it under Greenhouse's 2.5 MB parse limit.
3. The autofill check. When a Workday or other application offers to fill fields from your resume, go through every job and education entry before continuing. Fix wrong dates or merged title and company lines by hand.
4. The keyword check. Compare your resume against the job post, manually or with a matching tool, and add missing terms only where they're true.
5. The human check. Ask one person in your field to read it for 30 seconds and tell you what role they think you do. If they can't, rewrite the top third.

FAQ

Do ATS systems reject 75% of resumes?

There's no published evidence for that. The figure is widely traced to a 2012 marketing claim with no methodology. Automatic rejection usually comes from knockout questions that the employer configures, not from the system judging your resume's design.

Should I send my IT resume as a PDF or a Word file?

Follow the job post. If it doesn't specify, a text-based PDF or DOCX is the usual safe choice; Greenhouse, for example, accepts both. What matters is that the text can be selected and copied, and the file is small.

Can an ATS read two-column resumes?

Often badly. Greenhouse's documentation lists columned layouts, tables and text boxes among the causes of failed parses. Your original file may still be attached for recruiters, but a broken parse can hide you from keyword searches. A single column avoids the problem.

Do AI match scores decide who gets interviewed?

Generally they rank rather than decide. Greenhouse's Talent Matching produces scores that recruiters view, and in the Enhancv interviews only 2 of 25 recruiters said they used scores as a firm cutoff. Relevant, specific content is still the best way to score well.

How many keywords should an IT resume have?

There's no magic count. Cover the required skills from the post that you genuinely have, each once in Skills and once in context in a bullet point.

If you're an IT professional in Bangladesh and want a practitioner's read of your resume before a big application, you can get in touch here.